Increased AChR clusters – effect of pre- or postsynaptic PLS3V5 action?

Since no improvement in survival and motoric ability were detected in PLS3V5 overexpressing SMA mice, morphological analysis was performed on muscle sections to investigate whether PLS3V5 overexpression affects NMJ morphology. Thereupon, it was found that the size of postsynaptic AChR clusters was highly significantly increased in the Gastrocnemius muscle of P10 SMA + PLS3V5 and HET + PLS3V5 animals when compared to SMA and HET mice (Figure 37). At early time points of mouse development and still preceding axon arrival, patches of AChR spontaneously assemble along the midline of muscles in a process termed AChR-prepatterning (Lin et al., 2001, Yang et al., 2001). Upon nerve arrival, synaptogenic factors are released that further trigger and refine AChR clustering to sites of nerve innervation. In this context, the most important presynaptic factor known to trigger AChR clustering is the proteoglycan and MuSK receptor ligand Agrin (Wu et al., 2010). Since in the initial experiments PLS3V5 was ubiquitously expressed in SMA + PLS3V5 and HET + PLS3V5 animals, the question was raised whether PLS3V5 at pre- or postsynaptic sites accounts for the increase in endplate size of P10 animals.

Noteworthy, addressing this question was of particular interest since a presynaptic effect of PLS3V5 expression on AChR clustering would point at improved presynaptic signaling in transgenic mice. Initial support for the view of a presynaptic contribution of PLS3V5 overexpression on AChR cluster size came from a time course experiment: Here, it was found that the endplate size between SMA, SMA + PLS3V5, HET and HET + PLS3V5 animals did not significantly differ at the early time point P1 (Figure 39). Importantly, endplate size differences between PLS3V5 expressing and control mice were first seen around P4.

However, the CAG promoter driving PLS3V5 expression is ubiquitously active already very early during development, also in the muscle (Sakai and Miyazaki, 1997). Therefore, in case of a muscle specific effect of PLS3V5 on AChR cluster size, one would expect an effect of PLS3V5 on AChR clustering already during the AChR prepatterning process, just before first nerve terminals arrive. In line with that, an increase in endplate size should also be detectable at P1. Around P4, significant morphological changes take place in the context of mouse NMJ maturation. Principally, these changes comprise synapse elimination (axonal pruning) and concomitant conversion of the uniform patch-like shape of AChR cluster into a perforated pretzel-like appearance (Kariya et al., 2008, Murray et al., 2008). In this context, postnatal endplate maturation is known to be dependent on excitation from motor nerve terminals (Misgeld et al., 2002). The fact that endplate size increase in PLS3V5 transgenic mice falls together with motor axon rearrangement processes further suggests a presynaptic role for PLS3V5 in AChR clustering. At this point it should be remembered that PLS3V5 overexpression indeed results in significant changes of presynaptic organization, such as delayed axonal pruning as well as hyperinnervation of endplates. Therefore, these effects of PLS3V5 on presynaptic organization might well account for the observed increase in endplate size.

To further analyze the influence of presynaptic presence of PLS3V5 on endplate size, PLS3V5 was motor neuron specifically overexpressed in PLS3V5^fl_st/wt;Hb9-Cre^tg/wt mice. In heterozygous Hb9-Cre^tg/wt mice, Cre is located in the endogenous Hb9 locus, resulting in heterozygous knockout of Hb9 (Arber et al., 1999). Moreover, homozygous Hb9-Cre^tg/tg mice are even perinatally lethal and display a complete loss of motor neuron axons (Yang et al., 2001). Furthermore, AChR prepatterning has been shown to be clearly disturbed in Hb9-Cre^tg/tg mice with abnormal scattered distribution of AChR clusters across the diaphragm muscle at the early time point E12-E18.5. Although postsynaptic prepatterning defects of homozygous Hb9-Cre^tg/tg are long known, until now no study has ever focused on later time points in living heterozygous Hb9-Cre^tg/wt mice. In the present study, it was found that also Hb9-Cre^tg/wt mice exhibit AChR clustering defects by means of reduced endplate size in Gastrocnemius muscle as compared to controls at P10. This finding highlights the essential role of Hb9 for normal motor neuron development (Yang et al., 2001). Furthermore, these

findings underline the importance of normal motor neuron development and nerve innervation for proper AChR clustering. Most strikingly, endplate size was restored to normal level in PLS3V5^fl_st/wt;Hb9-Cre^tg/wt mice. As muscular PLS3V5 expression can be excluded in PLS3V5^fl_st/wt;Hb9-Cre^tg/wt animals, it seems plausible that presynaptic changes must account for the observed increase in endplate size. Since PLS3V5 overexpression led to an amelioration of the endplate size defects not only in SMA but also in Hb9-Cre mice, these findings together point at a general PLS3V5 neuroprotective effect, which might in turn impact on endplate size.

Nevertheless, despite clear evidence for presynaptic contribution of PLS3V5 overexpression to enhanced AChR clustering, an additional muscle specific effect of PLS3V5 from P4 time point on cannot be excluded. Many studies have focused on the relationship between the actin cytoskeletal organisation and AChR clustering in the past. In myotubes, F-actin filament formation clearly precedes Agrin induced AChR clustering (Dai et al., 2000).

Additionally, it has been shown in the same study, that AChR clusters did not form in the presence of Latrunculin, a toxin that sequesters globular (G)-actin and prevents F-actin assembly. From these experiments, it can be concluded that F-actin assembly is a prerequisite for the formation of Agrin induced AChR clusters. Furthermore, AChR clustering involves signaling by positive regulators of actin polymerization and filament growth, e.g.

Rho-family GTPases (regulate actin polymerization-driven processes) (Weston et al., 2003, Hall, 2005), p21-activated kinase 1 (effector of the GTPase Cdc42, which in turn regulates actin polymerization) (Luo et al., 2002) or geranylgeranyl transferase (enhances the membrane association and activation of GTPases) (Luo et al., 2003). Additionally, cortactin, which is an activator of the Arp 2/3 complex, has been shown to be important for AChR clustering (Madhavan et al., 2009). Most interestingly, however, many F-actin-associated proteins with impact on filament organization have been found to colocalize with AChR clusters at the mature or developing NMJ, such as vinculin (links integrins to the actin cytoskeleton) or filamin as well as α-actinin (function in F-actin bundling and interconnection) (Bloch and Hall, 1983, Cartaud et al., 2011). α-actinin has been shown to inhibit the rate of F-actin depolymerization, thereby stabilizing F-F-actin (Cano et al., 1992). This observation is in line with findings for PLS3 which, when overexpressed in LB or HEK cells, increases the F-actin to G-F-actin ratio (Oprea et al., 2008). Although mechanics of PLS3-mediated F-F-actin stabilization have not yet been sufficiently studied, the findings together indicate that F-actin filament stabilization through inhibition of depolymerization might be a feature common to actin bundling proteins. Assuming that PLS3 exerts a stabilizing effect on F-actin and on the background that F-actin formation is essential for AChR clustering and maintenance, muscle specific PLS3 might hence be positively involved in endplate maturation. To examine possible muscle specific contribution of PLS3, it would be interesting to compare AChR

cluster size between motor neuron specifically and ubiquitously PLS3V5 expressing mice.

Any increase in endplate size in PLS3V5 ubiquitously expressing animals compared to motor neuron specifically expressing mice would then indicate muscle specific effects.